Online Deception in Social Media

Page created by Lloyd Rhodes
 
CONTINUE READING
Online Deception in Social Media
University of Kentucky
                                                                                                                          UKnowledge
Library and Information Science Faculty
                                                                                                            Library and Information Science
Publications

9-2014

Online Deception in Social Media
Michail Tsikerdekis
University of Kentucky, tsikerdekis@uky.edu

Sherali Zeadally
University of Kentucky, szeadally@uky.edu

Follow this and additional works at: http://uknowledge.uky.edu/slis_facpub
     Part of the Library and Information Science Commons

Repository Citation
Tsikerdekis, Michail and Zeadally, Sherali, "Online Deception in Social Media" (2014). Library and Information Science Faculty
Publications. Paper 12.
http://uknowledge.uky.edu/slis_facpub/12

This Article is brought to you for free and open access by the Library and Information Science at UKnowledge. It has been accepted for inclusion in
Library and Information Science Faculty Publications by an authorized administrator of UKnowledge. For more information, please contact
UKnowledge@lsv.uky.edu.
Online Deception in Social Media
Online Deception in Social Media

Abstract

The explosive growth of social media applications has revolutionized the way we interact with
one another. However, the emergence and use of this online environment has also created new
opportunities for deception. We present a brief comparison between traditional (i.e., offline)
deception and online deception with a focus on social media. Furthermore, we explore some of
the factors that can affect the difficulty in achieving deception in social media and we use a
deception model to classify different online deception techniques. We also discuss the ease of
deployment and success of these techniques. Finally, we highlight some challenges that social
media designers must address in the future to protect social media users from online deception.

1. Introduction

The rapid proliferation of Web-based technologies have revolutionized the way content is
generated and exchanged over the Internet leading to an explosive growth in social media
applications and services. Social media enable the creation and the exchange of user-
generated content and the design of a wide range of Internet-based applications. This growth
has been fueled not only by the increase in the number of services but also by the rapid rate of
their adoption by users. Between 2005 and 2013, we have witnessed a 64 percent increase in
the number of people using social media1. For instance, Twitter usage increased by 10 percent
in the period 2010-2013. A total of 1.2 billion users connect through Facebook and Twitter with
their accounts24. However, the ease of getting an account also makes it easier for individuals to
deceive others. Previous work on deception has found that people in general lie daily and
several past efforts have attempted to detect and understand deception20. Throughout history,
deception has been used in various contexts along with technology (Second World War, Trojan
War, etc.) to enhance an attacker's deceptive action(s). Social media provide new environments
and technologies for potential deceivers. There are frequent examples of people that have been
deceived through the use of social media and some with devastating consequences in their
personal lives.

In this paper, we consider deception as a deliberate act with the intent to mislead others while
the recipients are not made aware or expect that such an act is taking place and that the goal of
the deceiver is to transfer that false belief to the deceived ones2,9. This perspective on deception
becomes particularly relevant when examining social media services in which the boundary
between protecting one's privacy and deceiving others becomes blurry. Furthermore, we also
argue that these false beliefs are transferred through verbal and non-verbal communication14
and deception is measurable and identifiable through verbal (e.g., audio or text), non-verbal
(e.g., body movement) and physiological cues (such as heartbeat).

                                                 1
Online Deception in Social Media
One may argue that training and raising awareness such as the one provided to security
personnel7 may be an effective avenue for protecting users of their social media. However,
people who are trained to detect deception perform worse in detection accuracy than people
that do not17 and evidence of a “privacy paradox” point to individuals sharing detailed
information even though they are aware of privacy concerns26 making themselves more
vulnerable to attacks. To make things worse, social media, as a set of Internet-based
applications, is also broadly defined term with multiple categories that include virtual
environments that are vastly different from one another15,16.

This work aims to present the concept of deception and explores its use in social media in
particular. We focus on the motivations for deception in social media and we explore various
deception techniques that have been used recently and their impact on social media users.
Finally, we discuss some of the challenges that we need to address in the future in the area of
deception in social media. While detecting and preventing deception are important aspects that
relate to the topic of deception, understanding online deception and classifying techniques used
in social media is the first step in fighting it. Our future publications will explore aspects of online
deception detection and prevention because the strict length constraint imposed on this paper
does not permit us to do.

2. Online Deception

Nature favors deception as a mechanism for gaining a strategic advantage. For example,
viceroy butterflies deceive birds by looking alike with monarch butterflies (which have a bitter
taste) thereby ensuring their survival as long as there are not too many in a system 8. Similarly,
humans have been using deception in connection to a benign or hostile intent3. In warfare, Sun
Tzu29 argued that “all warfare is based on deception.”

Social media services can be classified based on social presence/media richness and self-
representation/self-disclosure16. Social presence can also be influenced by the intimacy and
immediacy of the medium in which the communication takes place while media richness
describes the amount of information that can be transmitted at a given point in time. Self-
representation determines the control that users have in representing themselves whereas self-
disclosure determines revealing one’s information whether willingly or unwillingly. Using the
aforementioned characteristics, a table was developed by Kaplan and Haenlein16 that included
the following social media: blogs, collaborative projects (e.g., Wikipedia), social networking
sites (e.g., Facebook), content communities (e.g., Youtube), virtual social worlds (e.g., Second
Life) and virtual game worlds (e.g. World of Warcraft).We present an expanded classification of
social media (shown in Table 1) that also includes microblogging (e.g., Twitter) and social news
sites (e.g., Reddit). We place microblogging between blogs and social networking sites 15 and
social news sites above microblogging given their similarity to microblogging in terms of social
presence/media richness (limited content allowed to be communicated through the medium and
average immediacy as news come in) and low self-presentation/self-disclosure due to their
nature as content-oriented communities.

                                                   2
Social presence / Media richness
                                  Low                                                         High
                     Low     Collaborative    Social news sites        Content           Virtual game
Self-presentation              projects                              communities             worlds
/ Self-disclosure                Blogs         Microblogging       Social networking     Virtual social
                     High                                                 sites              worlds

Table 1: Social media classifications.

As shown in Table 1, social media which provide users with a lot of freedom for presenting
themselves are in the second row while social media that force users to adapt to certain roles or
have no option for disclosing parts of their identities are in the first row. Moreover, with an
increase in media richness and social presence, we note the transition from social media
offering just text for communication to rich media aimed to simulate the real world using verbal
and non-verbal messages as well as more immediacy in communications for virtual game
worlds and virtual social worlds. The differences between these types of social media services
affect how deception is implemented and its potential success.

In most social media platforms, most communications are text-based and are done
asynchronously. In such environments deceivers have a great advantage for altering content
which is a cheap way to deceive others. Zahavi31 pointed out the difference between
assessment signals that are reliable and hard to fake and, conventional signals that are easier
to fake. For example, in the real world if an elderly person wants to pass as a younger person,
he/she can dress younger or dye his/her hair and this will produce conventional signals.
However, it would be much harder to fake a driver’s license (an assessment signal). Social
media however provide an environment in which assessment signals are not required and are
not the norm making deception easier to achieve. For instance, gender switching online requires
often only a name change.

3. Difficulty in Achieving Online Deception

It is not surprising that the level of difficulty in achieving online deception is determined by
several factors associated with the deceiver, the social media service, the deceptive act and the
potential victim. These factors will determine how easy or difficult it is for a deceiver to engage in
online deception. High difficulty in achieving deception may deter potential deceivers while low
difficulty may be seen as an easy opportunity to deceive others. Figure 1 shows the various
entities (deceiver, social medium, victim) involved in online deception.

                                                  3
Figure 1: Entities involved in online deception.

3.1 The Deceiver

Several factors associated with the deceiver will determine the difficulty in achieving online
deception. These factors include expectations, goals, motivations, his/her relation to target and
target's degree of suspicion2. Expectation is a factor that determines the likelihood of success in
deception. More complex messages have a higher likelihood to succeed20. Goals and
motivations also determine the difficulty of deception. Goals are broader and long-term while
motivations consist of specific short-term objectives. They directly influence the choice and type
of a deceptive act. One motivation taxonomy, developed by Buller and Burgoon2, described
three different motivators for deception: a) instrumental where one can identify goal-oriented
deception such as lying on one’s resume on a social medium to get more job offers, b) relational
(also known as social capital) such as aiming to preserve social relations typical in online social
networks26, and c) identity such as preserving one’s reputation from shameful events on their
online profile. The aforementioned motivating factors in turn determine the cost (i.e., the level of
difficulty in achieving deception) of deception for a deceiver. For example, a deceiver motivated
to fake his or her identity will have to put more effort offline to succeed due to the presence of
signals that are much more difficult to fake rather than online where many of the identity-based
clues (gender, age, etc.) may take the form of just conventional signals (e.g., adding this
information to one’s profile page without verification). The difficulty in achieving deception is also
determined by the deceiver’s relation to a target. Familiarity with a target and its close social
network make it easier to gain trust and reduces the difficulty in achieving deception. Many
users assume that with technology comes enhanced security and are more relaxed in trusting
others online4. Further, the level of trust that individuals place on the deceiver, will also reduce
their degree of suspicion towards him or her thereby increasing their chances of being deceived.

                                                  4
The moral cost also increases the difficulty in achieving deception26. Morals can heavily
influence what deceivers consider as immoral in regards to withholding information or even
lying. In the real world the immediacy of interaction may make it much harder to deceive for
some individuals. In contrast, in the case of online environments distance and anonymity28
contribute to a loss of inhibition and therefore the moral cost is lower for deceivers.

3.2 Social Media

Social media require us to expand our perspective on how interactions are perceived between a
receiver and a sender during deception. For instance, the Interpersonal Deception Theory (IDT)
states that the interaction between a sender and a receiver is a game of iterative scanning and
adjustments to ensure deception success2.

Donath8 has suggested that if deception is prevalent in a system (e.g., Facebook community)
then the likelihood of success is reduced. It makes sense that the prevalence of deception in an
online community is a factor that also determines difficulty in achieving deception. Social media
services which encounter high volumes of deception will lead to communities that are more
suspicious. This will increase the number of failed deception attempts. Furthermore, by
increasing the target's suspicion the difficulty will increase deterring deceivers from the
community. Eventually some equilibrium may be reached. This rational however, suggests that
communities with low prevalence of deception will likely be more vulnerable to attacks since
suspicion will remain low for potential victims. Determining the prevalence of deception in a
community remains a challenging task.

Similarly the underlying software design of social media can also affect the degree of suspicion:
the level of perceived security for victims increases the chances of success for the deceiver11.
Software design can cause several assumptions to be made by users about the level of security
it provides. Some aspects of the design can make users more relaxed and less aware of
potential risks of being deceived. For example, individuals may falsely assume that profile
information on a social networking site is difficult to fake due to additional verification methods
such as email confirmation. Moreover, assurance and trust mechanisms for a system will
determine the level of trust between the sender and receiver11. Assurance mechanisms can
either reduce the probability of a successful deception or increase the penalties for deceivers 11.
High penalties will increase the difficulty for deceivers especially when the chances of being
caught are high. Assurance mechanisms are considered to be effective in certain contexts
where it is argued that the need for trust can be completely diminished. In social media,
assurance mechanisms are much harder to implement and as such penalties and the chances
of being caught may be or seem to be lower than those in offline settings and as such the cost
of deception is much lower. Finally, media richness is also a factor that determines the difficulty
in achieving deception. In this context, Galanxhi & Nah10 in their study of deception in
cyberspace found that deceivers experienced more stress when communicating with their
victims through text rather than an avatar-supported chat.

                                                5
3.3 Deceptive Action

Time constraints and the number of targets of an attack are factors that also determine the
difficulty in achieving online deception. The time available and the time required for a successful
attack are important especially in social media services where asynchronous communication
takes place. Moreover, the time required for the deception to be detected also determines the
effectiveness of the deception method used. For cases where deception must never be
discovered, the cost of implementation of the deception method may outweigh the benefits
especially when the penalties are high. The social space that deception is applied to and the
number of online user targets who are required to be deceived affect the level of difficulty in
implementing the deception method. A politician needing to deceive in his/her profile all of their
voters will face a more difficult challenge compared to a deceiver deceiving just one individual.
The type of deceptive act is also another important factor. Complex types of deceptive acts that
are guided by multiple objectives (e.g., identity and instrumental) are more difficult to achieve.

3.4 The Potential Victim

In traditional settings the target’s ability to detect deception may be a factor that determines the
difficulty in achieving deception. Online deception seems to be much harder to detect by users.
For example, in a study of Internet fraud using page-jacking techniques even experienced users
failed to detect inconsistencies present except for a select few who did detect deception
showing that it is not impossible11. Therefore, in social media a target's ability to detect
deception also depends to some extent on his or her Information Communication Technology
(ICT) literacy. Deceivers will have to evaluate their potential victim's ICT literacy. Individuals with
a high ICT literacy can have a significant advantage over casual Internet users and therefore a
cost and benefit analysis for a social engineering attack may be higher in this case.

4. Deception Techniques in the Social Media Environment

There are various techniques reported in the literature that can be used to deceive others in
social media environments and they include: bluffs, mimicry (e.g., mimicking a website), fakery
(e.g., forging a fake website), white lies, evasions, exaggerations, web page re-directions (e.g.,
misleading someone to a false profile page) and concealments (e.g., hiding information from
one’s profile)21. We use the communication model proposed by Madhusudan20 to classify
deception techniques for social media and evaluate their effectiveness in achieving deception.

4.1 Deception Model

The model (depicted on Figure 2) consists of a sender (S), the content or message (I), the
channel through which communication takes place (C), and the receiver (R). When a receiver’s
expected model (the S, I, C triangle) is different from the received model then deception has
occurred. This is also in line with Ekman's9 definition of deception who argues that a receiver

                                                  6
must not anticipate deception. By manipulating any of the S, I, C elements or combinations of
these deception is achieved. We present an overview of social media and identify factors (S, I,
C) and social media types where deception can be achieved with minimal efforts (i.e., low cost)
and at the same time results in a fairly high deception success rate (shown in Table 3). These
were identified from closely related literature on the topic and we present more information in
the following sections.

                                     Interaction without deception

                                        Interaction with deception

Figure 2: Interaction without/with deception.

        Social Media Type                   Low difficulty               High deception success
Blogs                            S, I                                S, I
Collaborative projects           I                                   -
Microblogging                    S, I                                S, I
Social news sites                S, I                                S, I
Social networking sites          S, I, C                             S, I, C
Content communities              I                                   I
Virtual social worlds            S, I, C                             S, I, C
Virtual game worlds              I, C                                C

Table 3: Manipulation of Sender’s identity information (S), Content (I), and Communication
channel (C) with low difficulty and high deception success.

                                                    7
4.2 Content Deception

Manipulating content is presumably the most common way of deceiving others. In social media
this can be achieved by falsifying information. Social media that focus primarily on content such
as blogs, content communities, social news sites, and microblogging are highly susceptible to
such deception. Technology today allows us to manipulate multimedia files to an extraordinary
degree. Tampering with images23 is an efficient way to fake content such as representing that an
individual traveled all around the world in one’s photos by altering them and broadcast these
images using social media. Such strategy may help a deceiver to elevate their social status and
gain trust of a victim in order to obtain further information. In addition to videos and images, the
ease of manipulating content, that at times, is heavily based on just text allows for a low cost for
deception and a high probability of success because of various factors such as a low
information literacy of receivers (e.g., critically evaluating content), the lack of expectation for
verifiability and even accountability. In addition, social media that offer profile management for
their users such as social networking sites and virtual social worlds are also susceptible
especially in cases where the initiation of new relationships is advertised. A competent deceiver
in affective writing may have a substantial advantage on these types of social media.

In contrast, collaborative projects such as Wikipedia are less likely to be affected by this kind of
deception (i.e., manipulating I). The difficulty in achieving deception may seem low but the
chances of success (at least over the long-term) are also low. This is because the software
design of these types of social media where many-to-many communication is supported
enables many people to review the content. Examples of content deception can be seen in
Wikipedia where not only vandals (people who alter content with intent to deceive others) are
eventually detected but there are people who assume a role to fight them25. Furthermore,
assurance mechanisms such as the requirement for content validity (tracing content back to its
source) are built into the system to ensure that content deception becomes more visible.
Another example of content deception in these types of social media is with open source
software managed by multiple users where it is much harder to add malicious content and
succeed in deception because multiple individuals evaluate the code before it is being released.
Virtual game worlds also have a low probability for deception success because of the strongly
narrated elements such as having a specific role that forces a player to a specific line of actions.

4.3 Sender Deception

Sender deception is achieved by manipulating the sender’s identity information (S).
Impersonation is a common example which often results in identity deception. This deception
falls under the category of identity theft in identity deception30. In this case, the deceiver may
gain access to an identity and use it to obtain additional information from their peers such as
house address, date of birth, and cell number. The failure to authenticate the sender’s
credentials will lead to a successful deception. Social media’s designs that have in-built high
self-presentation and self-disclosure enable sender deception at a low cost. Blogs and
microblogging can lead to stolen identities because there are no control mechanisms to verify

                                                 8
new users or their associated names. However, the damage caused by deception with these
types of social media is also likely to remain low and long-term success is probably not
guaranteed. Owners of the identity may become aware of the theft or other individuals familiar
with that identity may start identifying behavioral cues that do not match with that identity. In the
case of social networking sites and virtual social worlds, the cost of deception increases
because, cognitively, an individual will have to satisfy behaviors that are appropriate to the
identity he or she impersonates. The benefits however seem to be much higher in a social
medium context because access to an individual's social network can lead to an enhanced
ability to gain people's trust within the network and obtain information from them. The target in
these cases may not necessarily be the individual whose identity is stolen but others within his
or her social network. With no control mechanisms in place for identifying a source, arguably,
unregistered individuals which do not have an account on a service may be more exposed than
registered users for the social media services described above.

Social media (such as collaborative projects or virtual game worlds) with low self-presentation
and self-disclosure are likely to be more protected in terms of identity theft. This can be partially
attributed to their intended function. Collaborative projects, content communities and virtual
game worlds are heavily task-based. A user, who wants to obtain access from the impersonated
identity's social network will have to perform just as well as the identity being impersonated in
tasks and “act the part.” The cost is likely to be high and the success of the deception low and
short-term.

A middle ground between content deception and sender deception involves manipulating
information associated with an identity. These deception attacks can be categorized as identity
concealment where part of the information for an original identity is concealed or altered and
identity forgery where a new identity is formed30. For example, people may attempt to fake some
of the information in their profiles in order to gain trust or represent themselves in a different
way. In customer social networking sites, people may conceal information in order to gain
advantages of different offers 5.

4.4 Communication Channel Deception

Manipulating the communication channel requires a higher technical skill level which increases
the cost of deception. Tampering with the communication channel includes modifying in-transit
messages, re-routing of traffic, eavesdropping, etc. Jamming communications have been used
in virtual game worlds. Podhradsky et al.22 argued that multiplayer games in consoles can be
hacked in order to provide access to a user's Internet Protocol address. Once the intruder has
access to the host, he/she can kick the player out and proceeds with an identity theft deception.
In this case the goal of a deceiver may not be to obtain information but to damage a victim's
reputation. It is worth pointing out that there is a fine line between an unintentional
disconnection and an intentional departure of a player in a video game. This line becomes
blurred when the player is on the losing side and suddenly leaves. As a result, the reliability and
reputation of this player are damaged by an invisible deceiver. An advantage that
communication channel deception has is the implicit assumption that people make that digital

                                                 9
technology is imperfect and things may not work as good as in real world. Non-verbal behavior14
such as body movement or patterns of speech can expose deceivers, however, through social
media a deceiver may introduce jitter or delays in their video or audio in order to conceal their
deception, effectively increasing the chances of success. Victims at the other end of the
connection will have a difficult time in differentiating between an unreliable/slow connection and
the deceptive act being performed.

Since channel deception often involves technology all social media services may be susceptible
to an attack, especially those that use a similar set of technologies or architectures. Services
that have a higher reliance on their client applications will be more prone to attacks while those
that rely on server applications will probably be safer. Services with high media richness tend to
rely a lot on client software as is the case with virtual social worlds and virtual game worlds.
Deception, by exploiting communication channels, is common in such services13. Server-side
applications such as social networking sites or content communities are less prone to channel
deception because exploits rely on vulnerabilities of web browsers and web servers which are
generally hardened and made more secure.

The cost of this deception is quite high. However, the likelihood of success is also high
especially for well-orchestrated attacks.

4.5 Hybrid Deception Techniques

Hybrid deception techniques involve the manipulation of multiple elements (S, I, C) in the SIC
model described earlier and can be more effective in launching deception attacks. The
relationships between S, I, and C, as described by Madhusudan20, produce a consistent view for
a receiver. If one element of the SIC model shows a slightly different behavior, this may give
clues about an inconsistent relationship between two elements (e.g., S and I). For example, a
message received and signed by one's relative may lose its credibility if the source information
of the message does not match with that of the relative.

Various hybrid deception techniques that manipulate the content and the sender’s information
have been reported in the literature. These include examples such as forgery20, phishing,
identity forgery, web forgery11, email fraud. These techniques are highly effective in social media
such as social networking sites, virtual social worlds, microblogging and blogs, which highlight
user identities and provide a one-to-one or one-to-many communications. These online
deception attacks have not only been demonstrated to be effective but their consequences can
lead to disastrous consequences including the loss of life. In reality, the boy was actually the
mother of a former friend who used deception to gain her trust and later sent cruel and hurtful
messages to the girl. A service initially designed for people who want to initiate new
relationships and the lack of verifying both parties led to a devastating outcome. Online
deception can also have financial consequences as is the case with web forgery (e.g., creating
a website that represents a fake business). Web forgery involves manipulating the sender’s
information and the content. Web forgery becomes relevant for social media services due to the
increasing trend of including user-developed applications or widgets in many of these services.

                                                10
Even after an internal review mechanism that is effective in detecting malicious software,
vulnerabilities may still be unexpectedly present in these social media applications.

5. Challenges and Opportunities

The cost associated with successful deception in social media environments open up several
challenges which include: a) the lack of a standard, unified theory and methods in deception
detection for online contexts, b) the lack of universal or context-specific and computationally
efficient methods for deception detection for large online environments, and, c) the lack of effort
in deception prevention by social media developers.

5.1 Lack of a Standard Unified Theory and Methods for Online Deception

Currently, several theories for both online (e.g., phishing emails) and offline environments (e.g.,
employment interviews) have been used and proposed for detecting deception including
Management Obfuscation Hypothesis (MOH), Information Manipulation Theory (IMT),
Interpersonal Deception Theory (IDT), Four Factor Theory (FFT) and Leakage Theory (LT) 14.
These theories focus on detecting leakage cues that a deceiver gives away or strategic
decisions made by the deceiver which will reveal deceptive intentions. The main drawback with
these techniques is that they rely on a set of verbal and non-verbal cues that may not all apply
to the online world. For example, the presence of non-verbals in some social media requires us
to rethink what indicators can be used to measure them because they are not likely to exist
online in the forms that they exist in the physical world. A shift in focus is required for online
deception. Steps in that direction have been made with video blob analysis of hands and
movements for detecting movements that are too fast for the eye to detect (100% multiple state
classification accuracy but with a limited sample of only 5 interviews)19, detection of image
manipulation by detecting inconsistencies in compression artifacts (30% - 100% depending on
type of image, compression and tampering method)23, machine learning detection using audio
and transcribed text to identify patterns that signal deception because of deviations from a
baseline (66.4% accuracy when baseline is at 60.2%)12, and computerized voice stress analysis
to identify variations in an individual’s speech patterns (56.8% - 92.8% accuracy depending on
context)6. Furthermore, a promising aspect in social media is the fact that most of the verbal
cues are text-based. Methods of using verbal deception detection have been used to
successfully identify identity deception using techniques such as similarity analysis of profile
information (80.4% - 98.6% accuracy)30, similarity analysis along with natural language
processing to identify identity deception through writing patterns (68.8% accuracy)25, cross-
referencing information between a social network and anonymized social networks that contain
the nodes present in the first network to evaluate the trustworthiness of social network profile
attributes (40% - 80% recall depending on metric and technique when baseline recall at 20%)5,
and natural language processing to identify text features that betray deceptive emails (75.4%
accuracy)27. These techniques show that there are options available for addressing issues of
deception online. However, these aforementioned techniques cannot be directly applied to

                                                11
address all types of online deception for all types social media because: a) there is a large
variation among social media in terms of design and the type and amount of information that is
allowed to be exchanged between users, and, b) it is difficult to determine the context in which
optimum accuracy will be achieved for each solution. Put simply, the field lacks a cohesive
framework that captures the interdependencies and interactions among different detection
methods, types of deception and types of social media.

5.2 Computational Efficiency

The techniques that are currently used in deception detection are highly context-specific and
many of them cannot be applied to the online social media environment. Some of the most
popular detection deception methods dealing with verbal communication include Content-Based
Criteria Analysis (CBCA), Scientific Content Analysis (SCAN) and Reality Monitoring (RM) 14.
The applicability and effectiveness of these detection deception methods to social media is
unclear. Methods dealing with verbal cues such as video analysis may be computationally
inefficient19. Similarly, methods that aim to detect sender deception (identity deception) and use
similarity analyses to match identities may be feasible for small datasets but a comparison of all
records with one another results in a computational time complexity O(N2). In some contexts
where the profile information is available and text comparison is possible for features on a
profile, the time complexity can be reduced to O(w’N) using an adaptive sorted neighborhood
method30. The method sorts a list of records based on profile features and then moves through
the records using a window (w) that compares just the records within that window in order to
find duplicates. The adaptive method shortens the window (w’) by finding the first (if any)
duplicate record in a window and then ignores all further comparisons within that window (hence
w’ < w) which drastically increases the efficiency of the algorithm (1.3 million records parsed in
6.5 minutes). Similarity analyses are the ones that can most likely produce the highest
overheads especially in social media where datasets tend to be large. Scalability becomes an
issue for large datasets which will require more efficient approaches. For such cases, perhaps
techniques such as the Expectancy Violations Theory (EVT) which looks for deviations from a
baseline19 may be an efficient way of filtering suspect cases for further examination. This is a
computationally cheaper alternative that can be applied to both cases of sender and content
deception. For example, comparing deviations from a normal user baseline will require parsing
a database just once leading to a complexity of O(N). Finally, methods used in deception
detection in social media need to take into account social context features (such as friends and
family of an individual) which have been found to increase the accuracy of detection of
deception18. The downside to this is that social network analyses (SNA) tend to become
dramatically more expensive as networks grow. Simple SNA metrics such as betweeness
centrality becomes overwhelmingly difficult to compute as networks grow (O(N3)) where N is the
number of nodes and more advanced statistical methods such as exponential random graph
models which make use of Markov chain Monte Carlo algorithms become costly to compute.
Put simply, the potential for using these newly available social data is there, however,
computational efficiency needs to be addressed for large social networks. On a positive note, a

                                               12
new recent trend online is the formation of small social networking sites5 and communities for
which deception detection methods can become more computationally feasible.

5.3 Deception Prevention

Social media application designers need to make an effort to address the issue deception in
social media environments in the future. For example, Wikipedia’s policy requires information
added to articles to be cited back to its source and has ensured that many baseless arguments
are exposed to readers. Other social media services need to address the issue of identity
verification. For example, paradoxically, individuals who do not have an account today in the
popular social networking site, Facebook, are more likely to fall victims of identity theft (for
sensitive information) along with their real-life friends. The issue is that friends and other users
become wary in the presence of duplicate accounts especially when one has been active by the
original owner of an identity. On the other hand, when a deceiver registers an identity that did
not exist on a social media service before, users will be more likely to assume that the genuine
owner just joined the service. In an attempt to increase their user base, social media services
along with their easy registration and access features expose unsuspected individuals world-
wide to online deception. An effort to standardize the user registration and verifying users’
credentials needs to be investigated in the future.

6. Conclusion

The social media space keeps evolving and continues to be extended with a diverse set of tools
and technologies that deceivers can use. While the physical distance that separates the
deceiver and the victim may seem large, the damage that can be done is far from negligible.
Individuals, organizations and governments are at risk. Understanding how online deception
works through social media and future technologies remains a significant challenge. To address
this challenge we need to design social media applications with various rules and norms that
our traditional physical space does not have. Our desire for innovation has resulted in various
online social media designs that we do not yet fully understand and their vulnerabilities are
currently being exploited in various ways by various attackers including those involved with
deception attacks. It is time we start thinking about how to design social interaction in social
media environment to safeguard and protect social media users from the unforeseen
consequences of online deception.

Acknowledgements

We would like to express our gratitude to Elisa Bertino and Moshe Vardi for their
encouragements and support throughout the preparation of this paper. We thank them for the
opportunity they gave us to revise and strengthen this contribution. We also thank the
anonymous reviewers for their valuable feedback and comments which help us to improve the
quality and presentation of the paper.

                                                13
References

1.    Brenner J, Smith A. 72% of Online Adults are Social Networking Site Users.; 2013:15.
      Available at: http://pewinternet.org/Reports/2013/social-networking-sites.aspx .

2.    Buller DB, Burgoon JK. Interpersonal Deception Theory. Commun Theory.
      1996;6(3):203–242.

3.    Burgoon J, Adkins M, Jensen JKML, et al. An Approach for Intent Identification by
      Building on Deception Detection. Syst Sci 2005 HICSS ’05 Proc 38th Annu Hawaii Int
      Conf. 2005:21a–21a.

4.    Castelfranchi C, Tan Y-H. The role of trust and deception in virtual societies. Syst Sci
      2001 Proc 34th Annu Hawaii Int Conf. 2001:8 pp.

5.    Dai C, Rao F-Y, Truta TM, Bertino E. Privacy-preserving assessment of social network
      data trustworthiness. Collab Comput Networking, Appl Work (CollaborateCom), 2012 8th
      Int Conf. 2012:97–106.

6.    Damphousse KR, Pointon L, Upchurch D, Moore RK. Assessing the validity of voice
      stress analysis tools in a jail setting.; 2007.

7.    Dando CJ, Bull R. Maximising Opportunities to Detect Verbal Deception: Training Police
      Officers to Interview Tactically. J Investig Psychol Offender Profiling. 2011;8(2):189–202.
      Available at: http://dx.doi.org/10.1002/jip.145.

8.    Donath JS. Identity and deception in the virtual community. In: Smith MA, Kollock P, eds.
      Communities in Cyberspace. Routledge; 1999.

9.    Ekman P. Deception, Lying, and Demeanor. In: Halpern DF, Voiskounsky AE, eds. States
      of Mind : American and Post-Soviet Perspectives on Contemporary Issues in Psychology:
      American and Post-Soviet Perspectives on Contemporary Issues in Psychology. Oxford
      University Press; 1997:93–105.

10.   Galanxhi H, Nah FF-H. Deception in cyberspace: A comparison of text-only vs. avatar-
      supported medium. Int J Hum Comput Stud. 2007;65(9):770–783.

11.   Grazioli S, Jarvenpaa SL. Perils of Internet fraud: an empirical investigation of deception
      and trust with experienced Internet consumers. Syst Man Cybern Part A Syst Humans,
      IEEE Trans. 2000;30(4):395–410.

12.   Hirschberg J, Benus S, Brenier JM, et al. Distinguishing deceptive from non-deceptive
      speech. In: Interspeech 2005. Proceedings of Eurospeech’05; 2005:1833–1836.

13.   Hoglund G, McGraw G. Exploiting Online Games: Cheating Massively Distributed
      Systems. First. Addison-Wesley Professional; 2007.

                                               14
14.   Humpherys SL, Moffitt KC, Burns MB, Burgoon JK, Felix WF. Identification of fraudulent
      financial statements using linguistic credibility analysis. Decis Support Syst.
      2011;50(3):585–594.

15.   Kaplan AM, Haenlein M. The early bird catches the news: Nine things you should know
      about micro-blogging. Bus Horiz. 2011;54(2):105–113.

16.   Kaplan AM, Haenlein M. Users of the world, unite! The challenges and opportunities of
      Social Media. Bus Horiz. 2010;53(1):59–68.

17.   Kassin S, Fong C. “I’m Innocent!”: Effects of Training on Judgments of Truth and
      Deception in the Interrogation Room. Law Hum Behav. 1999;23(5):499–516.

18.   Li J, Wang GA, Chen H. PRM-based identity matching using social context. In:
      Intelligence and Security Informatics, 2008. ISI 2008. IEEE International Conference on.;
      2008:150–155.

19.   Lu S, Tsechpenakis G, Metaxas DN, Jensen ML, Kruse J. Blob Analysis of the Head and
      Hands: A Method for Deception Detection. Syst Sci 2005 HICSS ’05 Proc 38th Annu
      Hawaii Int Conf. 2005:20c–20c.

20.   Madhusudan T. On a text-processing approach to facilitating autonomous deception
      detection. Syst Sci 2003 Proc 36th Annu Hawaii Int Conf. 2003:10 pp.

21.   Nunamaker Jr. JF. Detection of deception: collaboration systems and technology. Syst
      Sci 2004 Proc 37th Annu Hawaii Int Conf. 2004:1 pp.

22.   Podhradsky A, D’Ovidio R, Engebretson P, Casey C. Xbox 360 Hoaxes, Social
      Engineering, and Gamertag Exploits. In: System Sciences (HICSS), 2013 46th Hawaii
      International Conference on.; 2013:3239–3250.

23.   Popescu AC, Farid H. Exposing digital forgeries by detecting traces of resampling. Signal
      Process IEEE Trans. 2005;53(2):758–767.

24.   Shen X (Sherman). Security and privacy in mobile social network [Editor’s Note]. IEEE
      Netw. 2013;27(5):2–3.

25.   Solorio T, Hasan R, Mizan M. A Case Study of Sockpuppet Detection in Wikipedia. In:
      Farzindar A, Gamon M, Nagarajan M, Inkpen D, Danescu-Niculescu-Mizil C, eds.
      Proceedings of the Workshop on Language Analysis in Social Media. Stroudsburg, PA:
      The Association for Computational Linguistics; 2013:59–68.

26.   Squicciarini AC, Griffin C. An Informed Model of Personal Information Release in Social
      Networking Sites. In: Privacy, Security, Risk and Trust (PASSAT), 2012 International
      Conference on and 2012 International Confernece on Social Computing (SocialCom).;
      2012:636–645.

27.   Stone A. Natural-Language Processing for Intrusion Detection. Computer (Long Beach
      Calif). 2007;40(12):103–105.

                                              15
28.   Suler J. The Online Disinhibition Effect. CyberPsychology Behav. 2004;7(3):321–327.

29.   Tzu S. The Art of War. Translated by Samuel B. Griffith. New York: Oxford University;
      1963.

30.   Wang GA, Chen H, Xu JJ, Atabakhsh H. Automatically detecting criminal identity
      deception: an adaptive detection algorithm. Syst Man Cybern Part A Syst Humans, IEEE
      Trans. 2006;36(5):988–999.

31.   Zahavi A. The Fallacy of Conventional Signalling. Philos Trans R Soc London Ser B Biol
      Sci. 1993;340(1292):227–230.

Michail Tsikerdekis is an Assistant Professor in the College of Communication and Information
     at the University of Kentucky. His research interests include sociotechnical systems and
     social interaction design in social media especially in the domain of social networks and
     collaborative projects.

Sherali Zeadally is an Associate Professor in the College of Communication and Information at
      the University of Kentucky. He is a Fellow of the British Computer Society and a Fellow
      of the Institution of Engineering Technology, England. His research interests include
      computer/system/cyber security, computer networking (wired/wireless), mobile
      computing, energy-efficient networking, and performance evaluation of systems and
      networks.

© ACM, 2014. This is the author's version of the work. It is posted here by permission of ACM
for your personal use. Not for redistribution. The definitive version was published in
Communications of the ACM, Vol. 57, No. 9, (September 2014)
http://doi.acm.org/10.1145/2629612

                                              16
You can also read